Using Google-based metrics to measure the impact of the bandwidth open up at RENU
MetadataShow full item record
In March 2020, all academic institutions that are schools, universities and other institutions of learning that compromise about 85% of the RENU membership were closed and the country thereafter forced into an undefined lockdown. This unprecedented lockdown put RENU in an uncharted position since most RENU members were closed with no specified reopening dates. This follows the structure of the Ugandan education system that heavily relies on physical learning and gathering of learners which was contradictory to the measures (SOPs) stated by the government to curb the spread of COVID 19. In response, RENU decided to open up its bandwidth to all its member institutions by removing any capacity limiters along the path. This meant that member institutions were not limited to their financially subscribed capacity but were now able to fully utilize as much capacity as they so needed. This response was meant to enable research institutions to work online seamlessly as they were critical in combating the novel COVID-19 and also to continue their very important work as they tackle major killer diseases in Uganda like malaria, HIV and natal deaths rates in the country. For education institutions, the response was meant to help them transition from only physical learning to blended learning which involves fewer physical interactions and more online learning. The response would remove any strains on the institutional bandwidth as more students would access the local institutional resources over the Internet, curb congestion and improve user experience to aid convenient learning. Some of the effects of the unlimited institutional bandwidth consumption included improved network performance and quality of service for the RENU member institutions which were backed up by statistics from Google ISP Portal, a tool that gave us an analytical 3D dimension look into Google based traffic usage right from the source of the traffic to the destination (end-user) factoring all players in the communication channel including the user device, i.e. the capability of the user machine( type of mobile phone, laptop or tablet) which all, in turn, have a contributing effect to performance and user experience. The tool uses metrics like Goodput, Retransmit rate and Application Round Trip Time to determine the performance of Google based traffic and also further breaks down the traffic according to the advertised BGP IP subnets which enables a deeper understanding of the general and per institution network performance. With such insightful information, RENU was able to understand the overall network performance as Google traffic constitutes about 60% of the overall RENU traffic. From the observations, poor institutional network performance was due to “bandwidth chocking” available bandwidth Vs required bandwidth. A correlation between high bandwidth capacity and improved network performance was also observed.